INDEX

Explanations

instances of the word "complaint" at varying activation levels

oai_token-act-pair · gpt-3.5-turbo

instances of the word "complaint" in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

lasses

-0.76

eton

-0.73

enta

-0.71

odes

-0.70

hered

-0.70

shock

-0.70

tera

-0.69

eele

-0.69

raham

-0.69

eston

-0.68

POSITIVE LOGITS

 alleging

1.11

 lodged

1.08

 filed

1.07

 complaints

1.03

 against

0.98

 complaint

0.97

 alleges

0.95

 levied

0.90

 leveled

0.89

 leve

0.83

Activations Density 0.058%

instances of the word "complaint" at varying activation levels

instances of the word "complaint" in various contexts

No Comments

No Known Activations